Cardiovascular Effects of Antiseizure Medications for Epilepsy

Abstract

Antiseizure medications (ASMs) are the primary treatment for epilepsy. However, adverse cardiac effects of ASMs can occur, related to their effects on lipid metabolism, raising ischemic heart disease risk; or specific actions on cardiac ion channels, increasing cardiac arrhythmia risk. Select ASMs, particularly enzyme inducers used at higher doses or for longer durations, can adversely affect lipids or cause metabolic changes, and thereby increase the risk for ischemic heart disease. These metabolic and potentially proarrhythmic actions may contribute to the increased cardiovascular morbidity and mortality that occur in epilepsy. Many ASMs block sodium channels or affect the QT interval, which can lead to proarrhythmia, particularly when used in combination with other medications or given to vulnerable populations. While ASMs are rarely reported to cause cardiac arrhythmias directly, population data raise concerns that cardiac arrhythmias and sudden cardiac death may be more common in epilepsy, and that sodium channel blocking ASMs in particular, might contribute. It is also possible that some cases of sudden cardiac death could be misclassified as sudden unexpected death in epilepsy (SUDEP), leading to an underestimation of the cardiovascular risk in this population. Cardiovascular risk factors, such as smoking and a sedentary lifestyle, are also associated with epilepsy, and should also be addressed. This summary is a narrative review of the literature, clarifies which ASMs tend to have more cardiovascular effects, and provides practical suggestions for medication management and monitoring from neurology and cardiology perspectives.

Key Points

Antiseizure medications can adversely affect lipid profiles and cause clinically significant cardiac arrhythmias, thereby increasing cardiovascular morbidity and mortality.

Assessment of cardiovascular risk by incorporating selected tests (e.g., electrocardiogram and lipid profiles), prior to and during drug treatment, can aid appropriate medication selection and dosage titration.

Cardiovascular risk factors such as increased tobacco use, physical inactivity, and obesity have also been associated with epilepsy and should be addressed.

Introduction

Epilepsy is associated with an increased risk of mortality, with standardized mortality ratios (SMRs) ranging between 1.8 and 5.09 [1–7]. Notably, cardiovascular disease and mortality are more common in epilepsy than in the general population [4, 8]. Seizures themselves can also directly cause cardiac arrhythmias [9–11]. While antiseizure medications (ASMs) are the mainstay of epilepsy treatment, it is important to recognize that ASMs might contribute to cardiovascular risk through their metabolic effects, thereby raising the risk of ischemic heart disease, or by increasing the risk for cardiac arrhythmias, which could lead to syncope and sudden death. The factors leading to increased cardiovascular risk are complex and include: comorbid cardiac and other medical conditions, epilepsy and seizure types, as well as underlying genetic predisposition. Since ASMs are integral to epilepsy treatment, it is important to understand how ASMs can affect cardiovascular health. This manuscript reviews the literature to examine how ASMs may affect this risk. This review provides a critical analysis and summary of the data, which are often contained in small case series and population-based studies, to help clinicians and researchers in the field. In addition, while ASMs might increase cardiovascular risk, it is important to recognize that epilepsy itself is also associated with seizure-associated, lifestyle, and comorbid medical conditions that independently contribute to cardiovascular disease risk, and those data are also included in this review.

There is a lack of data to guide clinicians on how to identify patients with epilepsy who are at higher risk for cardiac arrhythmias, or how to use electrocardiograms (EKGs) and other cardiovascular tests in managing ASMs in epilepsy treatment. With the use of societal recommendations and other data, this review provides practical suggestions to identify patients at risk, understand the use of EKG and other cardiovascular tests, and recognize when interventions are needed to prevent complications.

Literature Search

This narrative review includes data on the ASMs currently available in the USA. We reviewed English language papers published between 2004 and September 2024, although we also included older papers that included unique data or were needed to provide background to the current data. Keywords included individual ASMs, and one of the following: lipids, cardiovascular morbidity and mortality, arrhythmias, blood pressure, and EKG. We searched for cardiac risk prediction tools utilizing PubMed, supplemented by a Google search. Owing to limitations in the available data, we included case reports, case series, and population cohort studies focused on clinical studies; although some preclinical data are included when clinical data were limited. When there were multiple case reports or series reporting similar findings, we included representative papers that included the most data or the earliest date of report. Owing to the inclusion of case reports and series, statistical methods could not be applied, but papers that adequately described details supporting possible cause-and-effect or association of the clinical findings with the ASM were included. We did not include ASMs when there were no available data. Available data are summarized in a descriptive manner, and in tables and figures. Limitations and biases of the data are included in the descriptive portion of this manuscript.

Cardiac Disease and Mortality in Epilepsy

There is increased cardiovascular mortality in epilepsy [4, 8, 12], and a higher prevalence of cardiovascular disease in adults with (18.3%) than without (11.3%) epilepsy [13]. Nilsson et al. [4] found the standardized mortality ratio (SMR) for heart disease to be 2.6 [95% confidence interval (CI) 2.4–2.8] in a cohort of patients > 15 years of age who had been admitted with a diagnosis of epilepsy. In this study, they included ischemic heart disease, arrhythmias, as well as other diseases of the heart and lung in this category, but most patients had ischemic heart disease. In a study evaluating patients with epilepsy in rural China, many of whom had not been treated with ASMs, Ding et al. [8] found that the SMR for epilepsy was 3.9 (95% CI 3.8–3.9), compared with the general population. They found that cause-specific mortality for stroke (SMR 7.0, 95% CI 6.5–7.6) and myocardial infarction (SMR 10.7, 95% CI 5.6–95.3) were significantly higher in the epilepsy population. These data highlight the high risk of cardiovascular disease in people with epilepsy, even without ASM exposure.

Sudden cardiac death (due to cardiac arrhythmia) is the most common cause of sudden death in the general population, and a population study in the Netherlands found that epilepsy further increases this risk [14]. In that study, there were 12 people with epilepsy in a cohort of 1019 cases of sudden cardiac arrest, which translated into an approximately threefold increased risk [odds ratio (OR) of 2.9; 95% CI 1.1–8.0; p = 0.034] for sudden cardiac arrest in epilepsy over the general population, most pronounced in women and the young. A subsequent related study reported that it was actually the underlying cardiovascular disease, rather than epilepsy itself, that was responsible for cardiac arrest. In this study, patients with epilepsy who had ventricular tachycardia/ventricular fibrillation were more likely to have congenital/inherited heart disease, clinically relevant heart disease, and intellectual disability [15] than an epilepsy control group. It is possible that intellectual disability may have been related to the etiology of epilepsy (such as genetic factors), thus explaining this association. They did not find any differences in those taking ≥ 2 ASMs or QT prolonging medications, but the study was not powered to evaluate medications.

Using a large population database, Bardai et al. [16] found that ASMs were associated with increased risk for sudden cardiac death. They examined 926 sudden cardiac death cases, and 9832 age/sex matched controls, and found that sodium channel blocker (SCB) ASMs showed an increased risk (OR 2.8, 95% CI 1.1–7.2) while other ASMs did not. Notably, this study considered gabapentin to be a SCB drug, even though gabapentin appears to have a negligible effect on the sodium channel [17], and instead, has a stronger affinity to the voltage dependent calcium channel [17, 18]. Both carbamazepine (OR 3.2, 95% CI 1.1–9.2) and gabapentin (OR 5.7, 95% CI 1.2–27.9) increased risk for sudden cardiac death, but it is not clear that a SCB mechanism explains gabapentin's association. This study also only included 14 cases with epilepsy, and the numbers of cases and controls using any specific ASM were low (e.g., a total of 10 cases and 26 controls were taking carbamazepine), thus significantly limiting the strength of the findings. In addition, some patients were taking ASMs without a clear history of epilepsy, thus raising questions about whether they were taking these medications for another indication that could independently raise cardiovascular risk (such as for diabetic neuropathy, instead of epilepsy). Of note, they found that active epilepsy also conferred a higher risk of sudden cardiac death than seizure freedom, raising concerns that ASMs might increase the risk for ictal arrhythmias, which are common in epilepsy [9], and might specifically increase the risk for lethal cardiac tachyarrhythmias. However, additional larger population data in epilepsy are needed to fully evaluate this association.

Epilepsy is associated with increased ischemic heart disease-related death, and possibly, sudden cardiac (arrhythmic) death. Detailed evaluation of the possible role of ASMs in ischemic heart disease and arrhythmia is explored in subsequent sections.

Blood Pressure

There are limited available data regarding the blood pressure effects of ASMs. Rapid infusion of benzodiazepines or phenytoin can cause an acute drop in blood pressure, but it is not clear that long-term outpatient use of ASMs affects blood pressure [19–21]. Indirectly, ASMs can increase the risk of obesity, and thereby also increase blood pressure [22]. High doses of gabapentin can cause hypotension, bradycardia, and left ventricular systolic dysfunction in animals, and some data suggest that long-term gabapentin use for diabetic neuropathy increases the risk of cardiovascular disease by inhibiting voltage-gated cardiac calcium channels [23, 24]. A trend toward increased diastolic blood pressure, as well as increased heart rate stability over time, occurred in patients who later died from sudden unexpected death in epilepsy (SUDEP) compared with a control epilepsy population, but no association with any specific ASM was found [21]. It is not clear that the long-term use of ASMs has significant direct blood pressure effects, though there is a lack of reported adverse events. This is a difficult area to evaluate since there are many confounding factors that can affect blood pressure, including age, comorbid medical conditions and medications, as well as daily and diurnal blood pressure variations. More data are needed to understand the possible effects of ASMs on blood pressure.

Ischemic Heart Disease

Janszky et al. [12] reported an increased risk for, and worse prognosis following, myocardial infarction (MI) in individuals with epilepsy, including individuals without prior cardiovascular disease or stroke, and noted that increased smoking in people with epilepsy likely contributed. Concordant with these data, the CDC found emphysema, a marker of more significant tobacco use, was more common in patients with (5.5%) than without (1.7%) epilepsy [13]. This is relevant since patients with chronic obstructive pulmonary disease (COPD) are more likely to have cardiovascular disease (OR 2.46, 95% CI 2.02–3.0) and hypertension (OR 1.33, 95% CI 1.13–1.56) [25] than the non-COPD population. Epilepsy is not only associated with an increased likelihood of smoking, but also with physical inactivity and obesity, as compared with the general population [26]. These associated lifestyle risk factors increase the cardiovascular risk in epilepsy, independently from ASM-associated risks. However, both indirect lipid data and population cohort data suggest that ASMs may also increase ischemic heart disease risk in epilepsy.

ASMs, particularly enzyme-inducing ASMs (EIASMs), may increase the risk of cardiovascular disease by adversely affecting lipid metabolism, inflammation (C-reactive protein), homocysteine levels, and the risk for obesity/metabolic syndrome [27–29]. Several large population studies have evaluated the possible association between EIASMs and cardiovascular disease, and most studies reported an increased risk of myocardial infarction (MI) and/or stroke/transient ischemic attack (TIA) with EIASMs (Table 1) [5, 30–33].

Table 1:

Cardiovascular disease in epilepsy and use of EIASMs: selected studies

Study/study design	Patients	Cardiovascular disease (CVD) in epilepsy	Notes
Janszky et al. 2009 [12] Population-based case control study	1799 acute myocardial infarction (AMI) 2339 matched controls	Increased OR (4.92, 95% CI 2.34–10.31) for MI in epilepsy Did not evaluate EIASM use risk 8-year follow-up	Patients with epilepsy had worse MI prognosis
Renoux et al. 2015 [30] Population-based cohort study	252,407 adults using ASMs (for epilepsy and other indications)	EIASM > 24 months increased risk of MI (RR 1.46,95% CI 1.02–1.85) Mean follow-up 3.7 years	Use of enzyme inhibiting ASM decreased risk of MI (RR 0.81,95% CI 0.66–1.00)
Chen et al. 2016 [5] Population-based cohort study	7461 newly treated epilepsy	Increased risk of MI in epilepsy (SIR 4.18, 95% CI 3.54–4.91) EIASM: no increased risk of MI 5-year follow up	EIASM use increased risk of stroke (RR 1.78, 95% CI 1.14–2.77)
Lee-Lane et al. 2021 [28] Retrospective matched cohort study	10,241 epilepsy 35,145 matched controls	Increased MI in epilepsy (HR 1.58, 95% CI 1.51–1.63) EIASM did not increase risk Mean follow up 6.1 years	ASMs: 31% EIASMs
Josephson et al. 2021 [31] Retrospective open cohort study > 18 years	31,479 epilepsy	EIASMs: dose dependent increased risk of ischemic heart disease/stroke/TIA (see text for details) Median follow-up 9 years	ASMs: 37% EIASMs Duration–response effect, especially after 10 years of use
Mayer et al. 2024 [32] Retrospective case-control cohort study	374,950 adult epilepsy Compared CVD outcomes with use of lamotrigine, carbamazepine, and sodium valproate	Carbamazepine and sodium valproate associated with higher CVD risk than lamotrigine (HR 1.390, 95% CI 1.160–1.665 and HR 1.264, 95% CI 1.050–1.521, respectively)	Valproate associated with higher 10-year all-cause mortality than carbamazepine (HR 1.226, 95% CI 1.017–1.478)
Li et al. 2024 [33] Prospective cohort study	431 epilepsy within a cohort of 27,230 adults 45–85 years baseline	Increased risk of new onset stroke, TIA, or MI in epilepsy (OR 2.2, 95% CI 1.48–3.27) 6-year follow-up	EIASM associated with increased risk (proportion risk 24.6%, 95% CI, 6.5–54.6% for strong EIASMs; 4%, 95% CI 0.8–11%, for weak EIASMs)

EIASM use is associated with increased cardiovascular disease in most studies (except Lee-Lane [28]). Increased duration of EIASM use is associated with increased risk of cardiovascular disease [30, 31]

Strong EIASMs: carbamazepine, phenytoin, phenobarbital, and primidone. Weak EIASMs: ox-carbazepine, eslicarbazepine, topiramate, and rufinamide

EIASM enzyme-inducing antiseizure medications, MI myocardial infarction, TIA transient ischemic attack, OR odds ratio, RR relative risk, HR hazard ratio, SIR standardized incidence ratio

Renoux et al. [30] evaluated > 250,000 adults using ASMs for various indications, and found that EIASM use for > 24 months increased the risk of incident MI as compared with randomly selected age and sex matched controls within the cohort. In contrast, Chen et al. [5] found that patients with epilepsy (and no prior history of vascular disease) had an increased risk of ischemic heart disease [standardized incidence ratio (SIR) 4.18, 95% CI 3.54–4.9] over the general population during a 5-year follow-up period. They did not find that EIASMs specifically increase this risk; however, EIASMs were associated with an increase in the risk of cerebrovascular disease [relative risk (RR) 1.78, 95% CI 1.14–2.77]. While the cause for this discrepancy is not clear, it is possible that the relatively low number of MIs (17) compared with cerebrovascular disease (177) may have limited their ability to fully evaluate the effect of EIASMs on ischemic heart disease. Another limitation of this study is that it included only patients who were admitted to public hospitals, which may have biased the population towards an older cohort, possibly with more severe epilepsy or other comorbidities at baseline.

Similarly, Lee-Lane et al. [28] found that while epilepsy increases cardiovascular risk [adjusted hazard ratio (HR) 1.58, 95% CI 1.51–1.63], EIASMs did not increase this risk over ASM treatment with non-EIASMs. Notably, they classified individuals as being on an EIASM if they had at least two prescriptions for an EIASM, so it is not clear how long patients were taking these medications, thus raising questions about the degree of EIASM exposure in these individuals. They also did not control their findings for lifestyle-related cardiovascular risk factors, which also limits the strength of their findings.

Josephson et al. [31] evaluated the important question of whether duration and dose of EIASM exposure may affect cardiovascular risk. This study evaluated cardiovascular outcomes in adults with epilepsy who were taking long-term EIASMs with those who were taking non-EIASMs. This study controlled for age, epilepsy duration, and medical comorbidities that increase cardiovascular risk, as well as lifestyle factors. The HR was higher for propensity-matched EIASM users as compared with non-EIASM users (HR 1.21, 95% CI 1.06–1.39), and they also found an increased cumulative hazard after 10 years of follow-up. When the daily dose of EIASM was doubled, this elevated the HR from 1.54 to 2.38, over a maximum follow-up of 25 years. A duration of at least 10 years of drug exposure, and a higher dose of EIASM, was associated with a greater risk of cardiovascular disease than non-EIASM use, and this remained true after controlling for other risk factors.

A somewhat different approach was undertaken by Mayer et al. [32], who performed a case-control cohort study evaluating adults without a history of cardiovascular disease prior to the diagnosis of epilepsy, who were taking carbamazepine, lamotrigine, or valproate. They found that carbamazepine and valproate use was associated with an increased risk of cardiovascular disease compared with lamotrigine (HR 1.390, 95% CI 1.160–1.665 and HR 1.264, 95% CI 1.050–1.521, respectively). However, since their database depended upon available coding data, it is possible that in some cases, cardiovascular events predated the epilepsy diagnosis, and the authors could not confirm how long patients were taking specific medications.

Another recent study evaluated adults between 45 and 85 years of age and found an increased risk for new onset stroke, TIA, and MI in patients with epilepsy than in the control population (OR 2.2, 95% CI 1.48–3.27). They also found that EIASMs were associated with an increased risk [33] (Table 1). In contrast, valproic acid (an enzyme inhibitor) use appeared to decrease the risk of MI and vascular disease in some studies, possibly through beneficial lipid lowering effects [30, 34, 35], although this was not always seen [32]. Supporting this dose–response effect, long-term ASM use (both EIASMs and enzyme inhibiting medications, but not lamotrigine) may also increase atherosclerosis, as measured by carotid intima media thickness [36, 37].

While there are mixed findings from population cohort studies, those studies which evaluated dose and duration of EIASM use, and controlled for possible confounding risk factors, suggest that increased cumulative exposure to EIASMs increases the risk for cardiovascular disease. Valproic acid might decrease the risk for MI through its enzyme inhibiting effect, but there are conflicting findings, suggesting that there could be other mechanisms involved, such as increased risk for obesity. Thus, more data are needed to evaluate this. Epilepsy-associated cardiovascular risk factors also contribute to the increased burden of ischemic heart disease.

ASM Effects on Lipids

EIASMs adversely affect lipid metabolism and might increase the risk of cardiovascular disease (Table 2). The cytochrome (CYP) p450 enzymes play an important role in both the synthesis and the degradation of cholesterol to bile acids [38]. In particular, the CYP51 enzyme is involved in cholesterol biosynthesis, raising concerns that CYP51 enzyme-inducing medications could increase cholesterol levels [34, 38]. Several ASMs are metabolized through the CYP P450 system.

Table 2.

ASM effects on lipids

ASM	Lipids	Inflammatory markers	Weight gain
Cannabadiol	NA	NA	Weight loss reported
Carbamazepine	↑ TC, LDL ↑↔ HDL ↔TG	↑ CRP, HC	+
Divalproex sodium	↓↔TC, LDL ↕HDL ↑ TG	↑HC	+
Eslicarbazepine	↑ TC, LDL, HDL but smaller effect than with carbamazepine	NA	+ [179]
Lacosamide	NCS	−	−
Lamotrigine	NCS	−	−
Levetiracetam	NCS/+LDL	+/−	−
Oxcarbazepine	+/-TC, TG, HDL	+/−	+/−
Perampanel	NA	NA	+ [180–182] [180, 181]
Phenobarbital	+TC, LDL	−	−
Phenytoin	+TC, TG, LDL, HDL	+HC	−
Topiramate	?- HDL	+HC	−
Zonisamide	No increase reported [183]	No increase reported [183]	Weight loss reported

TC total cholesterol, LDL low density lipid profile, HDL high density lipid profile, TG triglycerides, CRP C-reactive protein, HC homocysteine, − denotes no effect, + denotes a reported increase, NCS no clinically significant changes, NA data unavailable

Isojarvi et al. [39] reported that carbamazepine increased total cholesterol and HDL levels, and transiently elevated LDL and triglyceride levels. Subsequent studies reported that carbamazepine and phenytoin can increase total cholesterol, triglycerides, and LDL levels, though individual studies show variability and some inconsistencies [40, 41]. Carbamazepine can increase total cholesterol by an average of 20–25 mg/dL and also increase C-reactive protein [39, 42–45]. More recently, a study reported that carbamazepine increases cholesterol over 12 month follow-up, but that lacosamide does not [42].

Valproic acid has also been reported to increase the total cholesterol, triglycerides, and LDL, and reduce HDL levels, but there is variability across studies, with some reporting lower total cholesterol and LDL levels [34, 40, 46–48] or only transiently increased levels [49]. As an enzyme inhibitor, valproic acid might have a beneficial effect on lipids, but it is possible that valproic acid-associated weight gain or other mechanism could potentially counteract this effect, resulting in the variable lipid effects reported. Further data controlling for weight gain and other cardiac risk factors may help to clarify the effects of valproic acid on lipids.

There are limited data on oxcarbazepine and eslicarbazepine. In a cross-sectional study, Pylvanen et al. [50] found no significant differences in lipids in men taking oxcarbazepine compared with controls. However, subsequent studies reported adverse lipid effects (elevated triglycerides and/or total and LDL cholesterol) after 3–6 month use [51, 52]. Yis et al. [53] found no differences in lipid levels or carotid intima media thickness in 21 children taking oxcarbazepine versus healthy control children. Conversion from an EIASM to eslicarbazepine improved the lipid profile [54, 55], but eslicarbazepine treatment may still cause a modest increase in lipids compared with placebo [56]. Trinka et al. [57] found that lipids were more likely to increase after initiation of carbamazepine as compared with eslicarbazepine. Overall, it appears that eslicarbazepine may have a relatively weaker adverse effect on lipids than other EIASMs.

Data from 11 clinical trials in individuals treated with pregabalin for diabetic neuropathy showed no clinically significant changes in total cholesterol, LDL, and triglycerides, but a small but significant reduction in HDL compared with placebo [58].

In summary, while there are some inconsistencies, literature suggests that EIASMs may cause adverse lipid changes, which can be reversible after drug discontinuation. As noted in Sect. 5, EIASMs may increase ischemic heart disease risk, particularly with long-term use, and some data suggest that enzyme inhibiting ASMs may decrease this risk. While the association is indirect, the lipid effect data, taken together with the cardiovascular risk data, suggest that EIASMs may increase ischemic heart disease risk, owing to their adverse lipid effects.

ASM Effects on Oral Anticoagulants

EIASMs can also affect the levels of direct oral anticoagulants (DOAC), which are used to prevent atrial fibrillation-related stroke and treat venous thromboembolism. Perlman [59] found that the peak levels of apixaban and other DOACs were lower in patients on EIASMs than control patients. EIASM use was associated with a 6.26-fold increased odds (95% CI 2.19–17.90) for apixaban concentrations below the expected range. In addition, Ip et al. [60] found that concurrent use of a DOAC with a cytochrome p450/p-glycoprotein modulating ASM was associated with a 28% increased odds of ischemic stroke compared with DOAC users not taking these ASMs (adjusted HR 1.28; 95% CI 1.05–1.57; p = 0.017). These findings raise concern that EIASMs can reduce the efficacy of DOACs and increase the risk of thromboembolic disease. Other potential drug–drug interactions (e.g., EIASM may render lipid-lowering medications less effective) can also occur [45, 61, 62], which could also increase cardiovascular risk.

ASM Effects on the Electrocardiogram and Risk for Arrhythmia

People with epilepsy may have a higher risk of cardiac arrhythmias than those without epilepsy [63, 64]. In the Wang et al. study [64] evaluating individuals without pre-existing cardiovascular disease, 11% of the 2699 people with epilepsy and 7.9% of the 326,733 people without epilepsy had cardiac arrhythmias (p_log-rank < 0.001 for all types of arrhythmias) during a median follow-up of 12.51 years (IQR 11.66–13.24). In addition, they found that while cardiac arrhythmias were more common in people with epilepsy who were not taking ASMs, compared with people who did not have epilepsy (HR 1.28, 95% CI 1.04–1.58), the risk of cardiac arrhythmia was even higher for those individuals with epilepsy who were taking an ASM (HR 1.40, 95% CI 1.22–1.60).

While both epilepsy itself and ASMs could contribute to cardiac proarrhythmia, ASMs have specific mechanisms of action that could directly cause cardiac conduction abnormalities. Cardiac depolarization is caused predominantly by the influx of sodium ions into a cell, while repolarization is caused by the efflux of potassium ions. Blocking of voltage-gated sodium channels can therefore cause widening of the QRS complex; while blocking of potassium channels increases the action potential duration and therefore the QT interval [65]. Drug-induced QT prolongation is generally due to blocking the rapid-delayed rectifier potassium current (IKr), which can cause QT prolongation and Torsade de Pointes (TdP) polymorphic ventricular tachycardia. Some ASMs block IKr and might be pro-arrhythmic [66, 67]. Clinical (and some preclinical) cardiac data for individual ASMs are summarized below. When available, reported cases of arrhythmia are included for each ASM (see Tables 3 and 4 for summaries of EKG effects and arrhythmias reported for specific ASMs).

Table 3.

ASM effects on EKG

May prolong PR interval	May affect QT interval
Carbamazepine	Shortening of QT interval
Eslicarbazepine	Cenobamate
Lacosamide	Lamotrigine
Lamotrigine	Primidone
Phenytoin	Rufinamide
Pregabalin	Prolongation of QT interval
	Carbamazepine
	Lacosamide
	Levetiracetam
	Phenobarbital
	Phenytoin
	Pregabalin

No effect or very limited data reported for other ASMs: brivaracetam, cannabadiol, clobazam, clonazepam, ethosuximide, felbamate, fenfluramine, gabapentin, ganaxolone, oxcarbazepine, perampanel, tiagabine, stiripentol, topiramate, valproic acid, vigabatrin, and zonisamide

ASM antiseizure medication

Table 4.

ASM electrophysiologic effects: clinical reports (see text for details)

ASM	Case reports	Notes
Carbamazepine	Bradycardia, AVB
Cenobamate	–	Contraindicated in familial short QT syndrome
Divalproex sodium	–	IV infusion: no QT effect
Lacosamide	Afib/atrial flutter, AVB, VT, asystole	Dose dependent effects on EKG
Lamotrigine	–	Unclear if supratherapeutic levels may increase risk of arrhythmia
Levetiracetam	QTc prolongation	Inhibits IKr
Oxcarbazepine	Bradycardia, tachycardia, AVB, VF (in Brugada syndrome)
Phenobarbital	–
Phenytoin	IV: bradycardia, AVB, asystole Oral: no effects at toxic levels	Affects IKr; class 1 B antiarrhythmic
Pregabalin	QT prolongation, AVB, reported in neuropathy
Rufinamide	Normalization of QTc in long QT syndrome case	Contraindicated in familial short QT syndrome

AVB atrioventricular block, IV intravenous, Afib atrial fibrillation, VT ventricular tachycardia, VF ventricular fibrillation, Ikr rapid-delayed rectifier potassium current, QTc corrected QT interval. No reported arrhythmias available for ASMs not included in this table

Brivaracetam A phase I study evaluating brivaracetam found no specific effects on QT interval [68].

Cannabidiol In vitro data suggest that cannabidiol may inhibit select cardiac ion channels [69] and there are limited reports of bradycardia [70]; however, these reports have occurred in recreational use of marijuana or its derivatives, not pharmaceutical grade cannabidiol.

Carbamazepine, oxcarbazepine, eslicarbazepine Case reports show that carbamazepine, a SCB agent, can cause reversible bradyarrhythmias, including AV (atrio-ventricular) block and asystole [71, 72]. Long-term carbamazepine use did not affect corrected QT (QTc) intervals or heart rate variability in one study of 36 patients with epilepsy, compared with healthy volunteers [73]. AV block was reported in a healthy 17-year-old girl on oxcarbazepine, which resolved after discontinuation [74]. Ventricular fibrillation occurred in a 30-year-old man with Brugada syndrome [75]. This latter case highlights the importance of screening for underlying cardiac disease when initiating ASMs. Eslicarbazepine may increase mean PR intervals, but values were still within normal limits, and no significant QTc interval changes were seen in 67 healthy adult volunteers [76].

Cenobamate and rufinamide A phase I trial evaluating the EKG effects of cenobamate, a SCB, found no clinically significant effects on heart rate, or PR and QRS intervals [77]. This study did find a slight non-clinically significant, dose-related, shortening of the QTc interval. Similarly, rufinamide, which prolongs the inactive state of the sodium channels, may shorten the QTc interval [78, 79]. Thus, cenobamate and rufinamide are contraindicated in patients with short QT syndrome. Notably, there is a case report of an individual with long QT syndrome and Lennox–Gastaut syndrome, whose QTc interval normalized with rufinamide treatment, suggesting a potential benefit for individuals with long QT syndrome [80], but additional investigation is needed to determine whether this is clinically useful.

Clobazam Clobazam does not appear to increase the risk for cardiac arrhythmia or significantly affect the QT interval [81].

Gabapentin and Pregabalin Animal and human data suggest that gabapentin and pregabalin may affect cardiovascular function, possibly via their effects on the voltage-gated L-type calcium channels [23, 82]. There have been case reports of AV block, and animal and human data reporting prolongation of the QT interval, with pregabalin; although most individuals had significant underlying cardiovascular risk factors and/or renal dysfunction (causing impaired clearance) [82–85]. Importantly, these individuals were treated for neuropathy and had different cardiovascular risk factors, so the risk in epilepsy is not clear.

Lacosamide Lacosamide results in slow inactivation of voltage-gated sodium channels, specifically inhibition of cardiac sodium channel SCN5A, which may be responsible for arrhythmias reported on this medication [65, 86]. A phase III study of lacosamide in individuals with diabetic peripheral neuropathy reported AV block (0.5%) and atrial fibrillation (0.5%), while PR prolongation was reported in another phase III trial [87, 88]. Pooled trial data showed a small dose-dependent prolongation of the PR interval at a maximum therapeutic dosage of 400 mg/day, and also a slightly increased QTc at a supratherapeutic dosage of 600 mg/day [89]. In 2010, DeGiorgio reported a case of atrial flutter/fibrillation in a 37-year-old woman with refractory focal epilepsy and no significant cardiac risk factors, after a dosage increase to 600 mg/day [90]. Subsequently, there was a concerning report of ventricular tachycardia in a 49-year-old man taking 400 mg lacosamide daily, in addition to carbamazepine, lamotrigine, clonazepam, and valproate [88]. However, in a study evaluating lacosamide in 220 healthy volunteers, QTc intervals did not increase at dosages up to 800 mg/day, though there was a small dose-related increase in PR interval [91].

Yadav et al. [65] reviewed 17 published cases (with 2 cases of intentional overdose) of cardiac arrhythmias (including those noted above) associated with lacosamide. The cases had a median age of 48 years (range 3–95 years), and the majority had at least one cardiovascular risk factor, including hypertension (52.9%), hyperlipidemia (17.6%), and coronary artery disease (11.8%). Several arrhythmias were reported in that review, including ventricular tachycardia (29%), atrial fibrillation, complete heart block, and pulseless electrical activity. Most cases were resolved with specific treatment of the arrhythmia and/or lacosamide discontinuation, and individuals were also typically taking other SCB ASMs.

In another study evaluating 85 patients who had received IV (intravenous) lacosamide treatment, 28 (32.9%) developed at least one cardiovascular adverse event (19/78 had new-onset first-degree AV block and 7/85 developed hypotension), which were mild or subclinical in most cases and did not require specific antiarrhythmic or inotropic treatment [92]. The mean PR interval in this group also increased significantly from 169.3 ms to 184.5 ms after treatment. A subsequent study also evaluating IV lacosamide use, similarly found new-onset first-degree AV block (3%) and increased PR interval, as well as a mild decrease in diastolic blood pressure [93]. Higher doses, polypharmacy (particularly with other SCB agents), and the presence of cardiovascular comorbidities, may increase the risk for arrhythmias [65, 66, 88, 94]. In 714 children and neonates, cardiac adverse effects were rare, with IV lacosamide causing bradycardia in one individual, and prolonged QT interval in another person (who had pre-existing long QT syndrome) [95].

According to the 2023 revised medication package insert for lacosamide, there have been post-marketing reports of bradycardia, AV block, and ventricular tachycardia, which have rarely resulted in asystole, cardiac arrest, and death [96]. It also notes that most cases occurred in patients with underlying conditions that increased the risk for arrhythmias, or were on other medications that affect cardiac conduction, and have been reported with both oral and IV routes of administration. It recommends caution in patients with cardiac disease (e.g., conduction system disease, sodium channelopathy) and those taking other medications that could also increase the risk of arrhythmias.

Lamotrigine Lamotrigine, which inhibits voltage-gated sodium channels, can also inhibit the IKr [97]. In October 2020, the Food and Drug Administration (FDA) added a warning to “avoid Lamictal in patients with certain underlying cardiac disorders or arrhythmias,” on the basis of in vitro testing of lamotrigine, which showed that lamotrigine exhibits class IB antiarrhythmic activity. Through IKr blockade, it is possible that lamotrigine could increase the QT interval and may increase the risk of ventricular arrhythmias [98]. In humans, lamotrigine has been reported to cause small reductions in QTc [79, 99, 100] and prolongation of the PR interval [100]. However, in a systematic review of 26 studies including almost 25,000 individuals, there was no conclusive evidence supporting an increased risk for sudden death or specific EKG changes [101]. Similarly, another study reviewing eight randomized controlled trials, as well as nonrandomized observational studies and case reports, found no conclusive evidence that lamotrigine significantly increases cardiac risk [102]. Christensen et al. [103] evaluated almost 92,000 new users (median age 46 years) of lamotrigine and found no increased risk of cardiac disease.

More recently, Wang et al. [104] analyzed Danish registers, identifying individuals over 65 years old taking lamotrigine, and found that individuals with toxic levels (median plasma concentration of 21 mg/L) had a higher incidence of all-cause and cardiovascular mortality, compared with individuals with non-toxic levels (median plasma concentration of 9 mg/L). However, it should be noted that only 6% of the 7286 individuals taking lamotrigine had at least one plasma concentration, thus limiting the strength of the data. This raises questions about why levels were drawn in those individuals, since the subset is so small. There may be bias affecting the data, as levels were drawn in specific individuals, perhaps owing to questions of adverse effects or other clinical factors. Additional data, including larger numbers of individuals with levels evaluated on a routine basis (and not specifically owing to any clinical concern), would be helpful in further understanding whether toxicity could have contributed to their findings. Subsequently, Aboukaoud et al. [105] reviewed the FDA adverse event reporting system database and found that there was no increased reporting of cardiac arrest or arrhythmia for an epilepsy indication. However, a psychiatric indication was associated with a sixfold increased reporting risk for cardiac arrest compared to an epilepsy indication, but those individuals were also more likely to use concomitant medications that could affect cardiac conduction, or have overdoses, which may have explained the differences between cohorts. Biehl et al. [106] reviewed a cohort of 233 veterans (average age 64 years, 87% with an underlying cardiac condition and almost half taking another SCB medication) and did not find any definite cardiac complications or deaths related to lamotrigine use, despite assessing a higher than average cardiac risk group.

At this time, it is not clear that lamotrigine causes an increased risk of sudden death or specific EKG changes. Very limited data suggest that supratherapeutic levels may be harmful but further evaluation is needed. On the basis of the available data in 2021, a combined International League Against Epilepsy/American Epilepsy Society (ILAE/AES) task force recommended considering EKG screening in individuals > 60 years or with cardiovascular risk factors (including diabetes, hypertension, familial hypercholesterolemia, and smoking), and obtaining a cardiology consultation if any significant conduction abnormalities are present [98]. They also recommend considering a repeat EKG once the patient is near, or at the target dose, of lamotrigine in people with known cardiac disease, or significant risk factors, or when someone is on other SCB or other medications that may impair cardiac conduction.

Levetiracetam Levetiracetam has multiple mechanisms of action, including inhibition of Kv3.1 (a delayed-rectifier potassium channel) [107]. Some case reports suggest that levetiracetam causes QT interval prolongation [108–111]. Rarely, ventricular tachycardia has also been reported, but has been associated with electrolyte abnormalities in one case and long QT syndrome in another [110, 112].

Several studies report no significant changes in the PR or QTc intervals with levetiracetam [113, 114]. A population study found that 104,655 individuals taking levetiracetam were not more likely to have cardiac arrhythmia or sudden cardiac death than those taking oxcarbazepine [115]. The overall risk of significant cardiac arrhythmias appears low, but individuals with underlying genetic disorders, such as long QT syndrome, may be at greater risk.

Perampanel Clinical trial data on perampanel did not find prolongation of the QT interval [116].

Phenytoin Phenytoin is a class 1B antiarrhythmic medication [117], but in vitro studies show that phenytoin can also block IKr [66]. Case reports of serious cardiac arrhythmias and hypotension (and in rare cases, death) have been related to IV phenytoin owing to rapid infusion of the cardiotoxic diluent propylene glycol [71, 117, 118]. In contrast, oral phenytoin toxicity did not result in significant arrhythmias, ECG changes, or deaths, despite levels of up to 75 μg/mL [119, 120]. Oral phenytoin likely carries a low risk of inducing cardiac arrhythmias but caution should be used with IV phenytoin, where the maximum rate of infusion is 1–3 mg/kg/min or 50 mg per minute, whichever is slower (FDA labeling, phenytoin sodium injection).

Phenobarbital and primidone Data on phenobarbital are limited. Primidone, which is metabolized to phenobarbital, can cause QT shortening rather than prolongation [67]. In a post-stroke population, phenobarbital was associated with a longer QTc than in patients treated with levetiracetam or no ASM, but baseline values were unavailable for comparison, and thus, cause and effect are not clear [121].

Tiagabine Tiagabine does not appear to affect selected cardiac ion channels [122].

Topiramate No EKG abnormalities were noted after IV infusion or oral administration of topiramate in 12 healthy volunteers [123].

Valproic acid Valproic acid affects voltage-gated sodium channels and T-type calcium channels. One study reported no EKG abnormalities after 6 months of treatment with valproic acid [124]. Intravenous valproic acid does not appear to affect the QT interval [125, 126]. However, one population study [64] reported that valproic acid was associated with increased risk for cardiac arrhythmias (HR 1.55, 95% CI 1.20–2.00), possibly related to its SCB mechanism.

Vigabatrin In a phase I trial evaluating the effects of vigabatrin (a selective GABA-T inhibitor) on EKGs, there were no changes in the QTc interval [127].

Zonisamide Zonisamide, which blocks T-type calcium channels, might prolong the QT interval in overdose, although this has not been uniformly reported [128].

Genetic Contribution to Arrhythmia For individuals with known cardiac conditions, such as long QT syndrome, ASMs may increase the risk for arrhythmia. Auerbach et al. [129] reported an association between SCB ASM use and arrhythmia risk for those with LQTS2. In a cohort study evaluating individuals with epilepsy and heart failure, valproic acid was associated with higher hazard of mortality due to all causes, as well as heart failure, which may be related to its effects on calcium channels, changes in blood pressure, and increased angiotensin 2 production [130].

Summary of ASMs and Arrhythmias Epilepsy may increase the risk for arrhythmias, irrespective of ASM use [64]. This may in part be related to its association with underlying structural or ischemic heart disease, or genetic predisposition. In addition, it has been postulated that repeated seizures themselves may cause catecholamine surges and hypoxemia, potentially causing cardiac conduction system dysfunction and coronary vasculature abnormalities over time, a condition that has been termed the “epileptic heart,” which may increase arrhythmia risk [131].

ASMs may directly affect the PR or QTc interval (as noted in Table 3) and might cause arrhythmias in some cases, particularly when a patient has an underlying cardiac or genetic condition. Clinically significant proarrhythmic potential appears to be strongest for lacosamide, but there are case reports suggesting that other medications can also, in rare cases, cause significant arrhythmias, especially in individuals with underlying risk factors. In addition, population data suggest that carbamazepine and valproic use may be proarrhythmic [64]. However, there are no details regarding specific individual cases, thus cause and effect are difficult to establish. Higher doses may increase the risk for arrhythmias. Additional data are needed to confirm these findings. SCB may overall pose the greatest arrhythmia risk for individuals with epilepsy, especially if they are taking multiple SCB medications.

EKG Monitoring and Individual Risk Assessment

The role of routine EKG screening in epilepsy is not well-established. While early repolarization patterns, mild QTc prolongation, longer P wave durations, and other findings have been reported to occur more frequently in epilepsy than controls; abnormal findings are infrequent and of uncertain clinical significance, and association with any specific ASM has been hampered by small sample sizes [132, 133]. Older age, male sex, and polytherapy correlated with higher likelihood of EKG abnormalities, but these factors are also associated with abnormal EKGs in the general population [134, 135]. Overall, while EKG abnormalities seem more common in epilepsy, the incidence of clinically significant and actionable findings on a single routine 12 lead EKG appears to be low [136]. In addition, limited long-term EKG data in epilepsy report that ictal or interictal bradycardia/asystole and other arrhythmias may occur [10, 137, 138], but these have not been clearly associated with any specific ASM.

However, as noted previously, there are several case reports of ASM-induced arrhythmia, so it would be helpful to identify who may benefit from EKG testing. Unfortunately, there are few guidelines regarding EKG screening and monitoring [98]. With the dearth of recommendations for ASMs, data from monitoring recommendations for specific antiarrhythmic drugs (AADs), which are more potent IKr and sodium current (INa) blockers than ASMs, can provide some, though more extreme, reference points. Sotalol and dofetilide are commonly used Vaughn–Williams class 3 AADs that specifically block IKr and prolong the QT interval. They are contraindicated when the QTc > 450 ms. Drug initiation requires inpatient telemetry monitoring for 3 days (five doses) with dose reduction when the QTc increases by 15% or QTc > 500 ms [139–141]. The IKr effect of ASMs is weaker, and thus, the current data do not support such stringent testing. However, for rapid IV high dose ASM infusion in hospitalized patients, particularly if receiving polypharmacy with other potentially proarrhythmic drugs or SCB polypharmacy or with cardiac proarrhythmia risk factors, the use of serial EKGs and telemetry monitoring might prevent complications.

Flecainide is a Vaughn–Williams class 1 AAD with the strongest use-dependent block of INa, can prolong the PR interval/QRS complex (both on average about 25%), and can cause AV block. It is contraindicated in patients with coronary artery disease, systolic left ventricular dysfunction (ejection fraction < 40%), and significant conduction system disease (e.g., left bundle branch block). Discontinuation is recommended if the QRS duration increases by > 25% [142]. It therefore seems reasonable to avoid or consider alternatives to SCB ASMs when significant conduction system disease (particularly left bundle branch block) is present, and to consider discontinuation if QRS duration increases by 25%. However, it must be emphasized that the AADs have greater effects on cardiac conduction than ASMs, and thus, these are conservative recommendations.

Specific screening questions in the clinical history may help to identify those with somewhat higher risk of underlying cardiac structural or genetic disorders; including a family history of sudden cardiac death or unexplained drowning in a first-degree relative, SIDS, or repetitive pregnancy losses (since these may be associated with cardiac channelopathies); or neuromuscular conditions known to be associated with ventricular tachyarrhythmias [143]. In these cases, further cardiac evaluation would be helpful before initiating medications that are potentially proarrhythmic, or alternative ASMs without proarrhythmia risk, should be initiated.

In general, a screening EKG seems reasonable for patients with risk factors for cardiac arrhythmia (Table 5) starting a SCB (and possibly also IKr blocking) ASM. Slow ASM titration to the minimum effective dose, monitoring for side effects, judicious use of surveillance EKG monitoring, and working with cardiology may help prevent ASM-related arrhythmias. A QTc interval > 460 ms in women and > 450 ms in men are generally considered prolonged [144], and it may be best to avoid SCB ASMs or IKr blockers when QTc is near these levels, particularly in high-risk individuals [145]. There is an approximately 5–7% increase in the risk of torsades de pointes (TdP) for every 10 ms increase in QTc interval, and a QTc interval > 500 ms significantly increases risk for torsades de pointes (TdP) [141]. When QTc is > 500 ms, intervention is needed, including ASM reduction/discontinuation when applicable. When the QTc interval is slightly prolonged, the risks and benefits must be weighed carefully, and evaluation of other modifiable contributing factors and cardiology input are helpful. As noted previously, rufinamide and cenobamate are contraindicated in familial short QT syndrome. In addition, individuals with long QT syndrome should avoid ASMs that prolong the QTc interval. However, additional data are needed to support specific guidelines and parameters. We have summarized general recommendations based on the above cardiology data, as well as the recommendations from the ILAE/AES task force for lamotrigine, in Fig. 1.

Table 5.

Risk factors for cardiac arrhythmias—consider EKG

*Age > 60 years

*Structural heart disease (e.g., ischemic heart disease, cardiomyopathy, congestive heart disease)

Renal or hepatic dysfunction (which could affect drug metabolism/clearance)

EKG: *Cardiac conduction disease (left bundle branch block, sinus bradycardia, atrioventricular and sinoatrial block) *QTc > 450 ms (men), QTc > 460 ms (women); high risk for TdP > 500 ms

*Genetic conditions: long QT syndrome, short QT syndrome, Brugada syndrome

Untreated thyroid disease (more commonly hypothyroidism)

Cardiovascular risk factors: diabetes, hypertension, familial hypercholesterolemia, smoking

*Significant electrolyte abnormality: hypokalemia, hypomagnesemia, hypocalcemia (typically hospitalized patients)

*Co-administration of other potentially QT prolonging drugs (e.g., antiarrhythmics, macrolide antibiotics, antifungals, neuroleptics, tricyclic antidepressants)

*Co-administration of multiple sodium channel blocking agents

In part modified from Yap et al., French et al., and Khatib et al. [98, 184, 185].

TdP torsades de pointes polymorphic ventricular tachycardia, QTc corrected QT interval, ms milliseconds

*Denotes the strongest risk factors (recommend EKG screening when present even in isolation).

Fig. 1.

Arrhythmia risk assessment/recommendations.*Electrolyte abnormality: hypokalemia/hyperkalemia, hypomagnesemia, hypocalcemia; **Per the American Epilepsy Society task force recommendations on lamotrigine, consider a repeat EKG near or at the goal dose. Repeat EKG could also be considered for other sodium channel blockers. Modified from Yap et al., French et al., Tisdale et al., and Khatib et al. [98,145,184,185]. AV atrioventricular, BrS Brugada syndrome, EKG electrocardiogram, LBBB left bundle branch block, LQTS long QT syndrome, SCB sodium channel blocker

In light of the adverse metabolic effects that are associated with some ASMs, individual assessment for cardiovascular risk might be helpful in assessing the risk/benefit ratio in initiating or continuing specific ASMs. The American College of Cardiology and American Heart Association guidelines on the primary prevention of cardiovascular disease recommends lipid screening from the age of 20 years [146, 147]. The American Heart Association released a guideline on the assessment of individual cardiovascular risk and developed an online risk calculator (PREVENT^TM) [148, 149], which might also aid management decisions in this population, particularly when ASM-associated lipid elevation is a concern. This tool uses individual clinical profiles (including age, sex, blood pressure, BMI, and lipid values) for patients between 30 and 79 years to estimate their 10-year risk of developing atherosclerotic cardiovascular disease. We have summarized general recommendations regarding the use of ASMs, specifically EIASMs, owing to their more consistently reported adverse lipid effects, in relation to ischemic heart disease risk in Fig. 2.

Fig. 2.

Ischemic disease risk assessment/recommendations. CVD cardiovascular disease, DM diabetes mellitus, EIASM enzyme-inducing antiseizure medication, HTN hypertension.

Modified from Arnett et al. [147] and PREVENT Online Calculator [149]

Role of ASMs in Sudden Unexpected Death in Epilepsy (SUDEP)

While SUDEP is not typically associated with ventricular tachycardia, it is possible that some cases are arrhythmic in etiology, as these potentially lethal tachyarrhythmias have rarely been triggered by seizures, and rarely reported in SUDEP [150–152]. Investigators evaluating post-mortem data from 68 SUDEP individuals found that 13% had variants in KCNH2 and SCN5A genes, which have been previously reported in patients with long QT syndrome [153], raising the possibility that cardiac arrhythmia could have contributed to some deaths. SUDEP most frequently occurs after a convulsive seizure, which appears to trigger severe cardiorespiratory dysfunction and death. Some studies have documented an association between the administration of ASM polytherapy and increased risk for SUDEP [154–156]. Detailed discussion of SUDEP and ASMs is outside the scope of this article, but recent studies suggest that convulsive seizures confer greater SUDEP risk than any specific ASM or polytherapy [157, 158]. In addition, Sveinsson et al. [159] found that polytherapy appears to reduce SUDEP risk, and they did not identify any specific ASM with increased risk.

Some reports suggest that specific ASMs, such as carbamazepine [160–162] and lamotrigine [163, 164], might increase the risk of SUDEP. However, a subsequent study evaluating patients from 42 randomized clinical trials revealed no significant difference in the rate of SUDEP between lamotrigine users and non-users [165]. Another study noted a tenfold elevation in mortality during pregnancy among women with epilepsy, primarily attributed to SUDEP [160]. It is notable that a majority (64%) of these SUDEP cases involved individuals on lamotrigine treatment, levels of which typically decrease significantly during gestation. The study by Sveinsson et al. [166] found that lamotrigine, whether as monotherapy or polytherapy, was not associated with an elevated SUDEP risk in women. Nightscales et al. [167] found no association between sodium channel blocking medications, including specifically lamotrigine, and the risk of SUDEP. Vigilant monitoring and dose adjustments of lamotrigine during pregnancy might reduce the risk of SUDEP by reducing the likelihood of breakthrough seizures.

Fenfluramine and the Risk of Cardiac Valvulopathy

Fenfluramine, which affects the serotonin system, had previously been approved as an appetite suppressant but was withdrawn in 1997, after reports of cardiac valvulopathy and pulmonary artery hypertension. The average weight loss dosage was 40–60 mg/day (range 20–220 mg/day) and 113 case reports of valvulopathy were reported to the FDA, with the majority (77%) being symptomatic and 24% requiring cardiac surgery, resulting in three deaths [168, 169]. The risk for aortic and mitral regurgitation with fenfluramine use was associated with duration of treatment > 6 months, and reports to the FDA found a 32.8% prevalence of valvular disease based on echocardiographic surveys [168, 170]. In addition, the valvulopathy risk increased ninefold with an increase in dose from < 40 to ≥ 60 mg/day [169].

In contrast, lower doses are used in the treatment of epilepsy. Fenfluramine’s antiseizure effects were first reported in the 1980s, with seizure reduction in photosensitive epilepsy as well as Lennox Gastaut syndrome [169]. More recently, fenfluramine has been shown to be effective in the treatment of Dravet and Lennox Gastaut syndrome at lower doses than previously used for weight loss [171–174]. A study evaluating 232 patients with Dravet syndrome using a maximum dosage of 26 mg/day, for a median duration of treatment of 256 days, found no cases of valvular heart disease or pulmonary artery hypertension [175]. Randomized controlled trials have not reported any cases of pulmonary hypertension or clinically significant cardiovascular disease [169]. No cases of valvular heart disease occurred during a median 23.9 month treatment of Dravet syndrome with fenfluramine at a median dosage of 0.44 mg/kg/day (maximum dosage of 26 mg/day) [176]. Long-term outcomes in ten patients with Dravet syndrome treated for a mean of 16 years reported four patients with inconsistently reported slight thickening of one or more cardiac valves, and two patients with stable and clinically insignificant valve thickening [177]. However, 2024 EU package labeling was updated to include a report of one child with Dravet syndrome who developed pulmonary artery hypertension (on 10.12 mg/day) that resolved after discontinuation [178]. The data thus far suggest that the lower doses of fenfluramine used in epilepsy may be less likely to cause significant cardiac valvulopathy or pulmonary hypertension, but the data are still limited. In addition, there is still a paucity of data in adults, and it is possible that the risk in children and adults may differ.

Conclusions

Epilepsy is associated with increased cardiovascular morbidity and mortality compared with the general population. These are, in part, related to epilepsy-associated lifestyle factors, including increased tobacco use, sedentary lifestyle, and a higher rate of obesity, as well as the seizures themselves. While the reasons are likely to be multifactorial, ASMs might increase risk by their effects on lipid metabolism, atherosclerosis, and cardiac rhythm. SCB ASMs can alter cardiac conduction and repolarization, leading to drug proarrhythmia, particularly when used at higher doses, in combination with other potentially proarrhythmic medications, and in individuals with underlying heart disease. Non-SCB ASMs should be prioritized in individuals at high risk for arrhythmia or who have arrhythmogenic EKG abnormalities (see Table 5 and Fig. 1 for details). Appropriate selection of ASM requires assessment of each individual’s cardiovascular risk profile. Weight-neutral or weight-loss promoting medications may be preferred in obese patients. Non-EIASM use should be prioritized in patients with hyperlipidemia or coronary artery disease (see Fig. 2 for details). Available data do raise concerns regarding the role that ASMs might play in sudden death, which may at times be classified as sudden cardiac death, but in other cases might be classified as SUDEP. While the frequency of cardiac arrhythmia is higher in patients with epilepsy than in the general population, it is not clear how often ASMs directly contribute to arrhythmias. More data are needed to determine the true frequency of these events.

Future studies are needed to develop guidelines for ASM management to reduce cardiovascular morbidity, particularly in individuals with underlying cardiovascular disease. While some general suggestions are provided, more data are needed to establish clear EKG testing guidelines (including specific acceptable limits for PR and QTc values), as well as the timing, frequency, and duration of monitoring needed, especially in patients with heart disease.

Declarations

Funding

The authors did not receive support from any organization for the submitted work.

Conflicts of interest

M.N. has received support for clinical trial research activities from Eisai and UCB. R.H. has received fees for educational activities from Boston Scientific, Inc. J.H. has no financial interests.

Ethics approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and material

Not applicable.

Code availability

Not applicable.

Author contributions

M.N. conceived the idea for the article. M.N., J.H., and R.T. worked on literature research and data extraction. M.N. and R.T. performed data quality assessment and completed the manuscript. All authors reviewed and critically revised the final manuscript.